LING/C SC 581: Advanced Computational Linguistics Lecture Notes Feb 3 rd

Administrivia Homework 3 is out today No lecture next week Therefore due date for homework 3 will be the following Tuesday by midnight You are welcome to listen to Prof. Bryan Heidorn of the School of Information this afternoon in my seminar class (Douglass 211; 3:30pm–6pm)

Today's Topics Revisiting Colorless Green Ideas and the Penn Treebank (WSJ) tregex lecture Homework 3

Colorless green ideas examples – (1) colorless green ideas sleep furiously – (2) furiously sleep ideas green colorless Chomsky (1957): –... It is fair to assume that neither sentence (1) nor (2) (nor indeed any part of these sentences) has ever occurred in an English discourse. Hence, in any statistical model for grammaticalness, these sentences will be ruled out on identical grounds as equally `remote' from English. Yet (1), though nonsensical, is grammatical, while (2) is not. idea – (1) is syntactically valid, – (2) is word salad One piece of supporting evidence: (1) pronounced with normal intonation (2) pronounced like a list of words …

Colorless green ideas Sentences: – (1) colorless green ideas sleep furiously – (2) furiously sleep ideas green colorless Statistical Experiment (Pereira 2002) wiwi w i-1 bigram language model

Part-of-Speech (POS) Tag Sequence Chomsky's example: – colorless green ideas sleep furiously – JJ JJ NNS VBP RB(POS Tags) Similar but grammatical example: – revolutionary new ideas appear infrequently – JJ JJ NNS VBP RB LSLT pg. 146

How is it used? One million words of 1989 Wall Street Journal (WSJ) material nearly 50,000 sentences (25 sections) Training: (2–21) 39,832 sentences Evaluation: (23) 2,416 sentences Standard practice training sections 2–21 evaluation % Penn Treebank (PTB) Corpus

Stanford Parser Stanford Parser: – a probabilistic PS parser trained on the Penn Treebank

Stanford Parser Stanford Parser: a probabilistic PS parser trained on the Penn Treebank

Penn Treebank (PTB) Corpus: word frequencies:

Stanford Parser Structure of NPs: – colorless green ideas – revolutionary new ideas

An experiment examples – (1) colorless green ideas sleep furiously – (2) furiously sleep ideas green colorless Question: – Is (1) the most likely permutation of these particular five words?

Parsing Data All 5! (=120) permutations of – colorless green ideas sleep furiously.

Parsing Data The winning sentence was: ①furiously ideas sleep colorless green. after training on sections (approx. 40,000 sentences) sleep selects for ADJP object with 2 heads adverb (RB) furiously modifies noun

Parsing Data The next two highest scoring permutations were: ②Furiously green ideas sleep colorless. ③Green ideas sleep furiously colorless. sleep takes NP objectsleep takes ADJP object

Parsing Data (Pereira 2002) compared Chomsky’s original minimal pair: 23.colorless green ideas sleep furiously 36.furiously sleep ideas green colorless Ranked #23 and #36 respectively out of 120

Parsing Data But … graph (next slide) shows how arbitrary these rankings are when trained on randomly chosen sections covering 14K- 31K sentences Example: #36 furiously sleep ideas green colorless outranks #23 colorless green ideas sleep furiously (and the top 3) over much of the training space Example: Chomsky's original sentence #23 colorless green ideas sleep furiously outranks both the top 3 and #36 just briefly at one data point

Sentence Rank vs. Amount of Training Data Best three sentences

Sentence Rank vs. Amount of Training Data #23 colorless green ideas sleep furiously #36 furiously sleep ideas green colorless

Sentence Rank vs. Amount of Training Data #23 colorless green ideas sleep furiously #36 furiously sleep ideas green colorless

Break

tregex The best introduction to Tregex is the brief powerpoint tutorial for Tregex by Galen Andrew. – The_Wonderful_World_of_Tregex.ppt

Homework Discussion useful command line tool: –diff

Homework 3 Consider small clauses, e.g. – John considers himself smart – John considers [ S himself smart] – himself smart is a "small clause" (no verb) – small clause has a subject: himself – small clause has a predicate: smart

Homework 3 docs/prsguid1.pdf (page 252)

Homework 3 docs/prsguid1.pdf (page 252)

Homework 3 docs/prsguid1.pdf (page 252)

Homework 3 Double Object Examples: – verbs give and persuade may take two objects – give Calvin a comic book – [ VP [ VB give] [ NP [ NNP Calvin]] [ NP [ DT a] comic book]] – persuade Hobbes to eat Calvin – [ VP [ VB persuade] [ NP [ NNP Hobbes]] [ S [ NP *][ VP to eat [ NP [ NNP Calvin]]]]]

Homework 3 Wikipedia on small clauses and constituency testing:

Homework 3 Using tregex find all the small clauses in the WSJ corpus Question 1: How many are there in total? Question 2: Would you expect the NP-PRD type to be the most numerous? Is it? Question 3: List all the different types and their frequencies (in descending order) Show your work: – i.e. show the tregex search terms you used